The billion-dollar boundary. The most contested piece of financial infrastructure in 2026 isn’t a blockchain. It’s the ramp that connects one to a bank account.
Over the past eighteen months, Stripe, Visa, Mastercard, and a cluster of specialist providers have poured billions into buying and consolidating crypto onramps across the US and Europe, because new stablecoin laws on both continents have turned the fiat-to-crypto boundary into the layer everything else depends on.. Follow the acquisitions.
The buying spree is the clearest evidence. Stripe paid $1.1 billion for Bridge in October 2024, at the time the largest acquisition in crypto’s history, and Bridge’s transaction volume more than quadrupled in 2025 even as bitcoin fell. Mastercard followed with a deal for BVNK worth up to $1.8 billion in March 2026..
Visa has been running stablecoin settlement pilots with select partners on supported blockchains, and that program has reportedly grown to roughly $7 billion in annualized volume across cards live in 18 countries.. Retail behaviour tells the same story from the other direction. A first-time buyer typically starts with a bitcoin calculator to see how much coin a fixed sum of dollars or euros actually delivers after fees, then completes the purchase inside an app in a couple of minutes, never learning which provider ran the identity check, sourced the liquidity, or absorbed the chargeback risk underneath.
Price discovery and checkout have merged into one screen. When a category of transaction stops feeling like crypto and starts feeling like buying anything else online, it has become infrastructure.. Why the ramp beats the chain.
Why does the ramp matter more than the chain? Because the chain is the cheap part. Moving a token between two wallets costs fractions of a cent on Solana and settles in seconds.
Moving a euro from a bank account into that token requires card network access, KYC tooling, sanctions screening, fraud provisioning, local payment rails, and a licence in every market you touch.. Little of that is optional anymore. The GENIUS Act, signed in July 2025, made stablecoin issuance a federally licensed activity in the US, and the OCC’s 376-page proposed rule from February 2026 reads like a bank charter.
Chainalysis’s year-end regulatory review argues the law has also become an international benchmark, with final implementing rules due this month and the full regime in force by January 2027.. Compliance at that grade is expensive, slow to build, and hard to fork. A blockchain can be copied in an afternoon; a licensing footprint across 40 jurisdictions can’t.
That asymmetry helps explain why the moat in crypto appears to have migrated from protocols to ramps, and why the companies that already own payment compliance at scale are the ones writing the cheques.. Europe ran the experiment first. In Europe, MiCA’s authorization deadline passed on July 1 this year, and issuers without approval are being pushed out of EU-regulated markets entirely..
The European experience shows what that filtering does to a market. Tether never sought MiCA authorization, so exchanges serving the European Economic Area delisted USDT through 2024 and 2025. Circle got licensed early through Ireland and watched USDC’s European transaction volume jump 337% in the first half of 2025, while roughly fourteen authorized issuers now operate around twenty compliant stablecoins across the bloc.
On this evidence, regulation didn’t shrink the market so much as decide who keeps it.. What the volume actually says. The volume justifies the spend.
Adjusted stablecoin transaction volume reached roughly $4.5 trillion in the first quarter of 2026, and every dollar of it entered the system through an onramp at some point. Total supply now sits above $310 billion by DeFiLlama’s count. Velocity has roughly doubled since early 2024, which suggests the existing float is working harder, not just sitting in wallets..
Here’s the part the press releases tend to skip: most of that headline volume still isn’t payments. Strip out trading, treasury shuffling, and bot activity, and genuine stablecoin payments run near $390 billion a year, around 0.02% of global payment flows. Small.
But the growth curve is steep, B2B transfers dominate it, and the constraint on expansion is not demand. It’s conversion capacity at the fiat boundary.. The business under the widget.
That’s precisely what an onramp is: conversion capacity, industrialised. The provider takes a card payment that can be charged back, delivers an on-chain asset that can’t be recalled, and prices the gap between those two facts into its spread. The distance between the market rate a buyer sees when they run the numbers and the amount that lands in their wallet is where the entire business model lives.
It’s an unglamorous credit-and-fraud business wearing a Web3 badge, and it’s the same business card networks have run for fifty years. No wonder they recognised it first.. There’s a second revenue stream hiding in the plumbing.
Between the moment a user’s fiat lands and the moment the stablecoin settles, the provider sits on float, and in a world of Treasury-backed reserves that float earns real yield, often 4% to 7% on the stablecoins in transit. Most white-label ramp providers keep it. Contracts increasingly turn on that single question, because at scale the interest can matter more than the headline fee..
Consolidation, from Washington to Open USD. The banks are further behind, and Washington hasn’t helped. Federal implementation has been messy enough that stablecoin rulemaking stalled in February after a White House meeting ended without a deal, leaving issuers to build against a moving target while final rules stay pencilled in for this month.
Uncertainty of that kind punishes small operators and rewards anyone with a compliance department the size of a mid-tier bank.. Which is roughly how the market is consolidating. On June 30, a consortium called Open Standard, backed by Stripe, Visa, Mastercard, Coinbase, BlackRock, and more than 140 other firms, announced Open USD, a stablecoin structured to hand most reserve income back to participants.
Circle’s stock dropped 13% on the news. The message wasn’t subtle: the distribution owners intend to own the money too, and the ramps are their distribution.. Even Tether, which spent a decade thriving offshore, seems to have read the map the same way.
It launched USAT in January through federally chartered Anchorage Digital, a token built to satisfy the GENIUS Act and reach American users through compliant channels. When the largest issuer in the world restructures itself around US ramp access, the direction of travel is hard to argue with..
Ramps in the wild. Specialist onramp providers still matter, for now. Transak covers more than 160 countries, MoonPay over 150, and aggregators route transactions across dozens of providers to find the best conversion rate for each user’s region and payment method.
Their footprints in local payment rails, especially outside the US and EU, are genuinely hard to replicate.. They’re also showing up in places that have nothing to do with trading. The World Series of Poker started accepting tournament buy-ins over Solana through MoonPay this summer in Las Vegas.
A poker room isn’t a crypto business. It’s a cash-handling business that found a faster rail, and it needed a ramp, not a whitepaper, to use it.. The pattern repeats in gaming, remittances, creator payouts, and contractor payroll.
Each vertical adopts stablecoins the moment the conversion step disappears into the interface. Which suggests an uncomfortable ranking for the industry’s self-image: user demand for blockchains looks modest, but user demand for cheaper dollar movement is enormous, and ramps are where one gets converted into the other..
The cross-border reality check. The strongest case is still cross-border. Traditional remittance corridors can cost senders up to 20% of the transfer, a number that should embarrass everyone in payments.
Stablecoins collapse that cost, but only where a local off-ramp exists to turn the token back into pesos, naira, or rupiah. In corridor after corridor, the binding constraint tends to be the ramp, not the rail.. There’s a twist in the data worth sitting with.
Cross-border activity as a share of stablecoin payments has actually been falling, with intra-country transactions climbing to nearly three quarters of volume by early 2026. People aren’t just using digital dollars to cross borders. They’re using them at home, as everyday money, in economies where the local currency or the local banking system disappoints..
That reframes what an onramp actually is. It stops being a door into crypto and becomes a door into dollars, which is a far larger market and a far more political one. Every emerging-market regulator now has to decide whether frictionless dollar access via a widget is financial inclusion or currency substitution.
The IMF has flagged both readings, and both have a case.. What builders should ask before signing. For fintech teams evaluating a ramp provider, four questions matter more than the pitch deck.
First, where does the money sit between the user’s fiat payment and on-chain delivery, and who earns the yield on it? Second, what’s the compliance model: does the provider act as a licensed principal, making your users their customers, or do you own the KYC relationship in a B2B2C setup?
The answer decides your liability, your data ownership, and how much of your KYC flow you can customise.. Third, look past headline fees to settlement reliability and quote transparency. The share of transactions that settle within 24 hours, whether the amount quoted upfront matches the amount delivered, and the escalation path when either fails will shape your support costs more than a half-point fee difference.
Fourth, map the provider’s corridor coverage against where your users actually are, because fee structures range from 0.5% to 4.5% depending on payment method and country, and a provider that’s cheap in Europe can be expensive or absent in the corridors that matter to you. Switching later typically costs months of rebuilding..
The question worth asking out loud. My read: the onramp wars will probably end the way payment wars usually end, with three or four gatekeepers, a thin layer of regional specialists, and interchange-style economics dressed in new vocabulary. The technology underneath will be genuinely better.
Settlement in seconds, programmable money, 24/7 finality. The market structure on top may look strangely familiar.. None of this makes the build-out pointless.
Cheaper cross-border transfers, working dollar access in weak-currency economies, and payment rails that don’t close on weekends are real gains, and they exist because ramps industrialised the messy edge between banks and chains. Infrastructure doesn’t have to be romantic to be useful..
So treat the ramp as the strategic decision it has become. If you’re building on these rails, choose your provider the way you’d choose a banking partner: audit the licences, negotiate the float, and assume the contract will outlive the hype cycle. And keep asking the question the industry has stopped asking out loud.
Crypto was designed to route around the institutions that control money’s entry and exit points. Fifteen years later, its growth depends on entry and exit points controlled by Visa, Mastercard, and Stripe. If the ramps belong to the same companies that owned the old rails, what exactly did we route around?.
Featured image credit. Tags: cryptotrends
